从视频中逐帧保存Mediapipe面网格地标

Question

I want to know how to save the output of facial landmarks frame by frame from a video using Mediapipe.

Answer 1

您需要迭代multi_face_landmarks并将它们附加到列表中，然后您可以执行以下操作：

columns = []
for i in range(1, 469):
    columns.append('x_'+str(i))
    columns.append('y_'+str(i))

data = pd.DataFrame(training_data, columns=columns)
data['label'] = 'somelabel'
data.to_csv('Data/filename.csv', index=False)